From Rainforest Tea to Clinical Trials

The Kalata-Kalata Tea

In the late 1960s, Dr. Lorents Gran was serving as a physician with the Norwegian Red Cross in what was then the Democratic Republic of the Congo. During his posting, he observed something that would quietly preoccupy him for the next decade: local women in the region routinely drank a tea brewed from the leaves of a plant called Oldenlandia affinis (family Rubiaceae), known locally as "kalata-kalata," during labour. The tea appeared to dramatically accelerate and strengthen uterine contractions — a uterotonic effect that local midwives had relied upon for generations.

Gran was both intrigued and scientifically troubled. The tea was prepared by boiling the leaves — yet its biological activity survived intact. Any protein he knew of would be denatured and inactivated by prolonged boiling. Whatever was in this tea was not behaving like a normal protein. The active compound was surviving conditions that should have destroyed it.

He collected plant material, returned to Norway, and began a painstaking biochemical investigation that would eventually yield one of the most consequential discoveries in peptide science — though it would take decades for the wider scientific world to recognise what he had found.

First Isolation: Kalata B1 (1973)

Gran spent years working to isolate and characterise the active component of the kalata-kalata extract. In 1973 he published two landmark papers identifying what he named kalata B1 — named directly after the local plant name. He demonstrated it was a polypeptide of approximately 30 amino acids with potent uterotonic activity, capable of stimulating uterine contractions with a potency comparable to oxytocin.

The isolation itself was a significant technical achievement for the era. Gran used a series of precipitation, chromatographic, and biological assay steps to narrow down the active fraction — work that would have taken months with 1970s laboratory equipment.

But Gran could not explain why kalata B1 survived boiling, acid, or the digestive system. The analytical chemistry tools of 1973 — principally amino acid composition analysis and partial sequencing — could not reveal what we know today: that the peptide backbone was closed into a ring, and that ring was locked in place by three interlocking disulfide bonds creating a topological knot. The cyclic structure went undiscovered. Gran's papers were noted, filed, and largely forgotten by the scientific mainstream.

"The compound was heat-stable, acid-stable, and protease-resistant — properties entirely inconsistent with a normal polypeptide. We could not explain it."

The 20-Year Mystery

For over two decades, kalata B1 existed in the literature as an anomaly. A small peptide with inexplicable stability. Potent biological activity at low concentrations. A structure that resisted every attempt at unfolding. Occasional papers referenced Gran's work, but no major research group took up the challenge of resolving the structural mystery.

The tools needed to solve it were being developed in other fields. NMR spectroscopy — Nuclear Magnetic Resonance — was advancing rapidly throughout the 1980s, driven by applications in organic chemistry and structural biology. By the early 1990s, NMR had reached the point where small protein structures could be determined in solution at full three-dimensional resolution, without the need for protein crystals. This was exactly the capability needed to reveal kalata B1's secret.

The puzzle pieces were assembling themselves independently. Peptide chemists were becoming interested in cyclic peptides as drug scaffolds. Mass spectrometry was improving to the point where the molecular weight of a small peptide could be determined with precision sufficient to detect whether an N-terminus and C-terminus were present or absent — the telltale signature of cyclisation.

The NMR Breakthrough (1995)

In 1995, a research team including members of what would become Prof. David Craik's group at the University of Queensland determined the three-dimensional structure of kalata B1 using NMR spectroscopy. What they found was astonishing.

The peptide backbone was head-to-tail cyclic. The N-terminus and C-terminus were joined by a covalent peptide bond, creating an unbroken ring — a structural feature unprecedented in characterised plant peptides at the time. And within that ring, three disulfide bonds connected six cysteine residues in an interlocking arrangement that literally threaded through itself: the Cys III–VI bond passed through the ring formed by the Cys I–IV and Cys II–V bonds and the connecting backbone. A true topological knot, embedded within a cyclic molecule.

This was the Cyclic Cystine Knot (CCK) motif. In one structural revelation, the 22-year-old mystery of kalata B1's stability was solved. The molecule was indestructible not because of exotic chemistry, but because of topology. No enzyme could unfold it because unfolding would require breaking covalent bonds — not merely disrupting the non-covalent interactions (hydrogen bonds, hydrophobic contacts) that normally hold protein structures together.

Boiling disrupts non-covalent interactions. Acid does too. Most proteases work by attacking accessible, flexible regions of peptide backbone. The CCK scaffold had none of these vulnerabilities. Gran had been right all along — the compound simply could not be denatured by any means available to a biological system.

Naming the Family: Cyclotides (1999)

The NMR structure triggered a surge of interest. Researchers began searching for other plant peptides with similar properties. Plants from the Violaceae and Rubiaceae families were subjected to peptidomic analysis (mass-spectrometry-based peptide profiling), and additional cyclic cystine-knotted peptides were rapidly identified.

By 1999, it was clear that kalata B1 was not an isolated curiosity but the founding member of a large and structurally coherent protein family. The term "cyclotide" was formally coined — from cyclo (cyclic) and peptide — to name this class of head-to-tail cyclic, cystine-knotted plant mini-proteins.

Professor David Craik's group at the University of Queensland's Institute for Molecular Bioscience became the focal point of the emerging field, establishing the structural rules (the six-loop CCK framework), cataloguing biological activities, and beginning to explore the pharmaceutical potential of the family. Craik's group developed the systematic nomenclature still used today and created the first comprehensive cyclotide databases.

The First Cyclotide Gene (2001)

In 2001, the first cyclotide gene was fully cloned and characterised. The discovery answered a question that had been nagging structural biologists: how does a plant make a cyclic peptide? Cyclotides are ribosomally synthesised — produced by the plant's normal protein synthesis machinery from messenger RNA, not assembled by non-ribosomal peptide synthetase (NRPS) enzymes as many other cyclic natural products are.

The gene encodes a larger precursor protein with a signal peptide, an N-terminal prodomain, and one or more cyclotide domains each flanked by short conserved recognition sequences. The precursor is processed and cyclised post-translationally by specific enzymes in the plant's endoplasmic reticulum and vacuolar compartments.

This was significant for two reasons. First, it confirmed that cyclotide diversity (the dozens of variants within a single plant species) arises from sequence variation in the precursor gene — meaning the full range of cyclotides in a plant could in principle be discovered by sequencing its genome. Second, it opened the possibility of engineering cyclotides at the genetic level: if you could modify the gene sequence, you could modify the cyclotide loop sequences while retaining the stabilising CCK scaffold.

Butelase-1: The Enzyme That Makes the Ring (2014)

A central unresolved question persisted through the early 2000s: which enzyme actually performs the final cyclisation step — the formation of the head-to-tail peptide bond that closes the ring? Plant proteases from the asparaginyl endopeptidase (AEP) family were implicated, but the specific enzyme had not been isolated and characterised.

The answer came in 2014, published in Nature Chemical Biology, from Professor James Tam's group at Nanyang Technological University, Singapore. The enzyme responsible was butelase-1, an asparaginyl endopeptidase isolated from butterfly pea (Clitoria ternatea). Butelase-1 recognises a short C-terminal sequence on the cyclotide precursor, cleaves it, and simultaneously ligates the new C-terminus to the waiting N-terminus — closing the ring in a single enzymatic step.

The discovery was transformative for laboratory peptide chemistry. Butelase-1 was found to be the fastest peptide ligase known to science — capable of cyclising peptide substrates approximately 20,000 times faster than the previously standard sortase A enzyme used in laboratory cyclisation reactions. It required no ATP, worked in mild aqueous conditions, and tolerated a wide variety of peptide sequences. Almost overnight, butelase-1 became the tool of choice for producing cyclic peptides in the laboratory.

Global Discovery: 50,000+ Cyclotides Predicted

As analytical methods improved through the 2000s and 2010s, cyclotide discovery accelerated dramatically. Professor Christian Gruber at the Medical University of Vienna led an initiative known as the Global Cyclotide Adventure — an extensive multi-year survey of plant species across Tanzania, Brazil, Venezuela, Hawaii, and Europe. Gruber's team combined field collection, mass spectrometry-based peptidomics, and sequence database analysis to systematically map cyclotide diversity across the plant kingdom.

The results were startling in scale. Cyclotides were confirmed in the Violaceae, Fabaceae, Cucurbitaceae, and Poaceae families in addition to the original Rubiaceae. Individual species in the Violaceae — particularly the genus Viola — were found to produce dozens of distinct cyclotide variants simultaneously. Extrapolating from partial surveys, estimates now suggest that the Rubiaceae family alone contains more than 50,000 distinct cyclotide sequences.

Cyclotides may constitute one of the largest protein families in the entire plant kingdom — yet they remained essentially invisible to science until the NMR breakthrough of 1995, and largely undiscovered until systematic peptidomics surveys in the 2000s. It is one of the more remarkable examples of a vast class of natural products hiding in plain sight in common plants.

The First Commercial Product: Sero-X (2017)

The first commercial product based on cyclotide technology launched in Australia in 2017. Sero-X is a biopesticide developed by Innovate Ag in partnership with Professor Craik's UQ group. It is derived from a water extract of butterfly pea (Clitoria ternatea) — the same plant that yielded butelase-1 — and contains cyclotides as its active insecticidal components.

Sero-X is registered in Australia for use on cotton, macadamia, and vegetable crops against a range of sucking and chewing insect pests. Its regulatory profile is exceptional by pesticide standards: it has no withholding period (crops can be harvested immediately after application), no maximum residue limit (MRL) concerns in major export markets, and is classified as non-toxic to bees and beneficial insects at field application rates. It degrades rapidly in the environment and does not bioaccumulate.

The commercial success of Sero-X validated the agricultural application pathway and demonstrated that cyclotide-based products could navigate regulatory approval, be manufactured at scale from plant material, and perform competitively against synthetic pesticide alternatives.

Clinical Trials: The First Cyclotide Drug (Present)

The most clinically advanced cyclotide in medical development is [T20K]kalata B1 — a single amino acid substitution (threonine-20 to lysine) engineered into the original kalata B1 sequence. This modification eliminates the membrane-disrupting toxicity of wild-type kalata B1 while preserving the structural stability of the CCK scaffold and conferring a new immunomodulatory activity.

The molecule was developed by Professor Christian Gruber's group at MedUni Vienna, who demonstrated that T20K showed remarkable ability to modulate T-cell activity and halt multiple sclerosis progression in animal models, including the established experimental autoimmune encephalomyelitis (EAE) mouse model of MS.

T20K was subsequently licensed to Swedish biotech company Cyxone AB, which has advanced it into human clinical trials. Critically, T20K is administered orally — patients take it as a tablet or capsule, not by injection. This is possible precisely because the CCK scaffold protects the peptide from gastric and intestinal degradation, allowing sufficient oral bioavailability to achieve therapeutic blood levels.

T20K represents the proof-of-concept the field has been building toward since the 1995 NMR breakthrough: a cyclotide engineered for a specific therapeutic application, surviving oral administration, entering clinical testing in human patients. It validates the foundational hypothesis that the CCK scaffold can serve as a platform for orally bioavailable peptide drugs targeting a wide range of diseases.